Initial occurrence

A few days after his first recording, Aksel demonstrated his ability to distinguish other Estonian stem allom orphs in addition to the present indicative stem. The da-infinitive was the first Estonian m arked form to ap p e ar (at 2;9;7), although it took about 6 months until it could have been assum ed as acquired.

L ooking dow n the First O ccurred colum n in Table 4 (at the end) indicates that the largest num ber o f initial new Estonian form s consisted o f the т я - infinitive stem with a past marker (starting from the age o f 2;9;8). T he first past marker to appear w a s initially as the past co pula oli, follow ed 11 days later by an o th e r -i m arked stem. A fter 12 days the -is m arking appeared, fo llow ed 23 or 26 days later by the -s marking.

In general, person m arkers appeared later, with person mark­

ings in present tense preceding those in past tense. T h e exception ap peared to be the early ocurrence o f the singular 1st person m arker, w hich actually had a very long acquisition period o f either 7.5 or 15.4 months.

The wtf-infinitive presented a puzzle: being m arked on the ma- infinitive stem, its em ergence was expected close to the past m arkings, and it appeared very early. How ever, its early o c c u r ­

T. Salasoo ₁₉₇ rence seem ed form ulaic, and its productive use began long after the -s past m arker was first seen.

T h e rcwd-participle was used for the first time even before the -v past m arking appeared, w hereas the perfect forms requiring in addition to the participle also person m arking o f auxiliaries, appeared last. T h u s the initial occurrence o f the Estonian m arkings was in the sequence o f the ^/я-infinitive, past m ark­

ings, person m arkings in present tense, follow ed by person m arkings in past tense and perfect form s - all this supporting my initial hypothesis.

In English, A k s e l’s use in his first recording o f the English past auxiliary verb for phrasing a question and the irregular past participle (m arked by -n) may have been form ulaic at this stage.

The latter was suspected o f being part o f a com m on form ula “ all gone” . Productive use o f the past participle was assum ed about a month later at 2; 10; 1.

The -ing o f the continuous form, the irregular past stem s and the negative imperative forms began to be used early, by the time Aksel was aged 2; 10; 1.

T able 3 confirm s that after a 3 and a h a lf m onths interval other English forms needing auxiliaries: the conditional and negation in present tense appeared, as well as the only person m arking in English, the singular 3ld person. T he regular past m arking -ed emerged over 2 and a h a lf months after that (and 5 m onths after the occurrence o f the irregular past form), and over a m onth later the future (ne eding an auxiliary) and the infinitive (needing the particle ‘t o ’) w ere used.

Thus, the initial em ergence o f the English person m arking prior to the regular past m arking does not support my hypothesis.

6.2.5. A cquisition

All the 11 Estonian and 4 English suffixal verb m arkings, that w ere used during the observation, w ere also assum ed to have been acquired. Table 4 includes also the acquired forms produced with the aid o f auxiliaries etc.

] 98 Use o f initial Estonian and English verb form s by a bilingual child Although the few occurrences o f m arked English verbs does not provide absolute certainty, glancing dow n T able 4 indicates most strikingly that the English m arkings -ing and the -n o f the irregular past participle, as well as the irregular past and the negative im perative form s, w ere all acquired before any Estonian m arkings were.

Acquisition o f the English conditional and present negative preceded that o f the -ed o f the regular past tense, to be followed by the future and infinitive form s (Table 3 and Figure 2). The last English m arking to have been acquired w as the person m arker -s o f the singular 3 rd person in present tense, 11.8 months after its initial appe ara nce, thus providing confirm ation o f the expected sequence.

In Estonian, the -n u d m arker o f the past participle appeared to have been the first m arker acquired, by the time it was first sighted w hen Aksel was aged 2; 10; 18, to be followed shortly by the past tense m arkers -s, -i and -is, in that order, with acquisi­

tion periods ranging from 0 to 2.2 months. (A lthough there was som e am biguity about w h e th e r acqusition o f the -5 past marker took 3 days or o ver 7 months, the shorter period seemed more likely.)

Acquisition o f the -т а (possibly) and -da markers, assumed to involve only a single constituent, sim ilar to tense marking, was co m p leted after that o f past m arkers.

M arkings o f all the persons (except the plural 2nd person, w hich was not used at all) seem ed to have been acquired last, their acquisition periods being longer ( 4 .6 -1 5 .4 months).

Person m arking o f verbs marked for past tense was a diffe­

rent matter, however. Apart from the -d m arking o f the plural 3 rd person, which may have been acquired by the age o f 3;4;23 (when it was also used in place o f other persons), none o f the person m arkers appeared to have been acquired in past tense even by the end o f the observation, 10.9-11.3 m onths from their initial use (show n by a plus sign in Figure 2).

In perfect tense, the negative and the singular 2nd and 3rd person forms were acquired around the sam e time as person forms

T. Sal as oo 199 were, p roba bly being d ependent on the acquisition o f their auxiliary verb forms. This is dem onstrated by the perfect singular P 1 person form not having been acquired by the end o f the observation, 7.3 months after its initial appearance, since the I s' person -n m arking was probably acquired only by that session.

Thus, it appears, that in both languages, A k s e l’s acquisition o f person m arking took a longer tim e and it was com pleted later than that o f other verb markings. In Estonian, acquisition o f the combination o f past tense and person m arkings appeared to take even longer.

7. In conclusion

In a very brief glimpse o f how a child in A ustralia was learning simultaneously 2 languages: his m other tongue, Estonian and the local language, English, it has been shown:

• how Aksel at 2 years and 9 months, when his v o cabulary consisted o f w ords from almost all word classes, yet expressed mainly as unmarked word s t e t r ^ he w as already able to keep the 2 languages separate, according to his conversation partner;

• how he produced mixed utterances m ostly w hen he did not know a particular concept in the language he was using, yet hardly mixed his developing grammars;

• how, based on his verb data, over time, his gram m ars in both languages w ere developing from the simple to the more com plex via m arking and use o f other aids, such as auxiliaries, particles, etc.

More specifically, it seems that expressions o f verb m arking by this bilingual child support well the the im plicational hierarchy hypothesis, namely, that m arking which involves only a single constituent o f a phrase, i.e.,

• in English, the -ing o f the continuous form, the -n o f the irregular past participle and the -ed o f the regular past tense;

200 Use o f initial Estonian and English verb forms by a bilingual child

• in Estonian, the past m arkings -is, and -s, the -da o f the Jrt-infinitive and the -n u d o f the past participle (the т а-infi­

nitive rem aining am biguous and not providing any evidence);

• all these precede that o f m arking persons, that involves exc hange o f gram m atical inform ation betw e en two consti­

tuents.

R eferences

Jusczyck, P. 1997. Constraining the Search for Structure in the Input.

Paper at Language Acquisition, Knowledge Representation and Processing. GALA ‘97, The University of Edinburgh, April 4-6.

Kaplan, J., J. Bresnan 1982. Lexical-Functional Grammar: a formal system for grammatical representation. - J. Bresnan (ed.), The mental representation of grammatical relations. Cambridge, Massachusetts:

The MIT Press, 173-281.

Pye, C. 1992. Language loss among the Chilcotin. - International Journal of the Sociology of Language 93, 75-86.

Pienemann, M. 1985. Learnability and syllabus construction. - K. Hylten- stam and M. Pienemann (eds.), Modelling and assessing second language acquisition. Clevedon: Multilingual Matters, 23-76.

Pienemann, M. 1988. Psychological constraints on the teachability of languages. - W. Rutherford and M. S. Smith (eds.), Grammar and Second Language Teaching. New York: Newbury House, 85-106.

Pienemann, M. 1989. Is language teachable? Psycholinguistic experiments and hypotheses. - Applied Linguistics 1, 52-79.

Pienemann, M. 1998. Language Processing and Second Language Acquisition: Processability Theory. Amsterdam: John Benjamins Publishing Company.

Salasoo, T. 1995. Morfoloogiliste tunnuste esmakasutamine ühe lapse arenevas keeles (Initial use of morphological markers in the deve­

loping language o f a child). - Keel ja Kirjandus 4, 239-252.

Salasoo, T. 1996a. Observations in the natural acquistion of Estonian morphology: a mix-and-match o f stems and suffixes. - M. Martin and P. Muikku-Werner (eds.), Finnish and Estonian - New Target Languages. Jyväskylä, Finland: Centre for Applied Language Studies, University o f Jyväskylä, 86-1 15.

Salasoo, T. 1996b. Natural acquisition o f some morphosyntactic features of Estonian in native and non-native environments. Poster presented at AILA 96, Jyväskylä, Finland, 4-9 August 1996.